Mobile terminal positioning methods that are based on positions of Wireless Local Area Network (WLAN) access points suit particularly well for indoor positioning due to the global existence of WLAN connectivity in buildings. Moreover, such methods are easily scalable, because additional infrastructure is needed only on the server side in order to compute the location estimation, for instance in the form of servers providing radiomaps (e.g. maps indicating at which positions access points can be observed) and/or positioning services.
Indoor mobile positioning can for example be done based on WLAN Received Signal Strengths (RSS) collected at different floors inside the buildings. There are typically two stages in WLAN-based positioning:
First, the training phase or data collection stage, where the data is collected in the form of so-called fingerprints, which contain location estimate (e.g., if available, based on a Global Navigation Satellite System (GNSS), sensor-based, WLAN-based, manually inputted, etc.) and the measurements taken from the radio interface(s) at the respective location. The training can for example be a continuous background process, in which mobile terminals are continuously reporting measured data to the server or learn their internal offline radiomap.
Second, the estimation/positioning phase or data estimation phase, where the mobile terminal estimates its current location based on the data available from the training phase.
The measurements collected and stored in the training phase may for example contain (specifically in the WLAN case):                Signal strengths (for example, RSS index, physical Rx level in dBm ref 1 mW, etc.)        Basic Service Set IDs (BSSIDs) (e.g. Medium Access Control (MAC) addresses) of the WLAN access points observed and, possibly,                    Service Set IDs (SSIDs)            timing measurements (Round-Trip Time)                        
This measurement data gets uploaded/reported to the server or cloud, where algorithms are run to generate models of the WLAN access points for positioning purposes. Such models may be coverage areas, node positions, radio propagation models (e.g. path-loss models), etc. In the end, these models or parts of them are transferred back to the mobile terminals for use in position determination (terminal-based positioning). Alternatively, the models may be stored in a positioning server to which the mobile terminals connect as clients for position information (terminal-assisted positioning).